Try to attach a faster DAC-converter chip to an FPGA to receive and transmit RF

Author
Pinnei
Max Output Level: -90 dBFS
  • Total Posts : 3
  • Joined: 2017/02/23 02:29:54
  • Status: offline
2017/04/12 09:37:07 (permalink)

Try to attach a faster DAC-converter chip to an FPGA to receive and transmit RF

Hello, everyone


I'm currently trying to attach a rather fast DAC-converter chip to an FPGA to receive and transmit RF in the future, but getting the converter to run and attaching a signal generator and an oscilloscope for testing is my main goal right now.


I'm coming from the digital world. I did all kinds of digital circuits and used an AD converter for easy tasks with a microcontroller, but when it comes to analog high speed signals which are differential and sensitive to several factors like impedance and so on, I have basically no idea what I'm doing.


The chip I want to use for this project is the AD9862 was purchased from . It's rather old but they are not very expensive, easy to solder, and they have been used by Ettus Research in several models of their USRPs which I'm using as a reference platform. If you have a suggestion for a better chip, please tell me!


Now the main thing I'm concerned with is the whole analog domain. The AD9862 has 2 differential inputs which can be optionally buffered (which is what I should do, right?) and the datasheet says the input buffer has a constant impedance of 200 Ohm. Now what I want to do is, just take those two AD-channels out to an unbalanced SMA connector with 50 Ohm impedance to connect a signal generator or a radio frontend later. So, I need a Balun for that.


Ettus did that too. They have several daughterboards which you can attach to the baseboard to have different frontends connected to the AD-/DA-Converter. Now, if I look at the BasicRX Daughterboard which does exactly what I want, I see that they are using a Balun called ADT1-1WT. When I look that one up, the datasheet tells me, that it has an impedance of 75 Ohm. Isn't that totally wrong? I thought I need a 50 Ohm unbalanced to 200 Ohm balanced transformer.


Also the input is terminated with a 50 Ohm resistor and the output which goes, directly without any more components except a connector, to the AD (VINP_A/VINN_A and  is series terminated (right? Or is that a low pass filter with the 10pF capacitor? I read on a mailing list somewhere that the values for a low pass filter are wrong in this schematic BTW) with 50 Ohm. That doesn't match at all with the 200 Ohm input impedance of the AD input. It would be awesome if somebody could explain that to me! For me all of the values are totally off.


Also, what about the traces on a PCB? They also need to have the right impedance to prevent reflections and standing waves. So I need to match them, I guess? So the output of the balun should be differential traces with a differential impedance of 200 Ohm going to the AD input and on the other side of the balun, I need a 50 Ohm trace going to the SMA connector?


These are all things you only seem to learn at university if you take electrical engineering as a major and I took computer science and this is all just an amateurish hobby for me, so I'm kind of lost right now 

Any suggestions will be grateful!
#1

1 Reply Related Threads

    savageopera
    Max Output Level: -78 dBFS
    • Total Posts : 625
    • Joined: 2008/04/01 23:26:15
    • Status: offline
    Re: Try to attach a faster DAC-converter chip to an FPGA to receive and transmit RF 2017/04/13 06:42:14 (permalink)
    I just take the two prongy thingy and plug it in to that two hole dealy in the wall..........works most of the time.....unless the dog chewwed it......then it sparks............sorry pal, but this is way over my head.
    post edited by savageopera - 2017/04/14 04:05:36

    Sonar Artist, HP Laptop, AMD A8700 , 1T+250g, M-Audio Fast Track Ultra ,Roland 88 Hammer action, Roland AX-1, M-audio 88es, Arturia minilabII......When I was young I wanted to become a mad scientist. I achieved everything except the "scientist" part....http://www.soundclick.com/savageopera
    #2
    Jump to:
    © 2024 APG vNext Commercial Version 5.1